On 2/5/2021 10:02 AM, Magdalena Banwinkler wrote:

External Email - Use Caution

Dear Freesurfer experts,

We are investigating brain structure differences between 2 groups. We are using Gray Matter Volume and Cortical Surface Area as structural measures and want to look at specific ROIs. We want to include Sex, Age and TIV as covariates. However, we are unsure if we should correct for TIV when using Cortical Surface Area or only when using Gray Matter Volume. We would be very happy about any kind of help or literature regarding this question!

Yes, I would do it for both. If you're using version 7+, then you can add --etiv to mris_preproc and it will divide the area or volume by the etiv (but you can use it as a regressor instead). For thickness, you do not need to correct for etiv.

Additionally, we have calculated a whole brain GLM analysis and found significant differences between groups. After inclusion of covariates, no result remained significant. However, when using an ROI approach including the same covariates, the results strongly indicated between group differences. Does anyone have an idea on why this could be the case?

It might be that you have a mild effect across all the voxels that does not reach significance until you average all the voxels together

Thank you very much for your help!

Best,
Magdalena, Helena & Federica

_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer